A two-dimensional optical phased array, or beam former, can for example be used in an infra-red counter measures (IRCM) system, for directing a high-power beam of potentially multi-band light at the optical aperture of a threat to dazzle and jam its optical seekers. These phased arrays can improve the survivability of military and commercial platforms under attack from threat munitions and missiles that may be guided by a variety of electro-optic (EO) and infrared (IR) seeker types, such as semi-active laser (SAL) designator sensors and EO/IR imagers that sense one or multiple wavelength bands, with these seekers often sharing the same optical aperture. For at least these reasons, two-dimensional optical phased arrays are of interest to providers of military and commercial aircraft, as well as to any commercial developer of IRCM systems for military and civilian applications.
A two-dimensional optical phased array can also be used in compact laser radar systems. Such systems are used as altimeters for aircraft (including rotorcraft). Such systems also are envisioned for some automobiles. An advantage of a phased array is its large field of regard and fast beam steering, which means that one phased array could take the place of 4 to 10 or more small mechanical or micro-electro-mechanical (MEM) beam steerers.
Current approaches for mechanical beam steering use gimbals that are not conformal, have slow response for large slews, and are capable of engaging only one threat at a time. To be specific, the gimbal must protrude beyond the outer-mold-line of the platform, it has fairly slow slew rates of >100 millisecond for 120 degrees of slew, and it requires a separate optical aperture for each of the multiple beams, thus enlarging the size and reducing the speed of the gimbal further.
Until recent years, the state-of-the-art for non-mechanical bi-dimensional optical beam steering devices was a liquid crystal based optical phased array, such as described in the document by P. F. McManamon, T. A. Dorschner, D. C. Corkum, L. J. Friedman, D. S. Hobbs, M. K. O. Holz, S. Liberman, H. Nguyen, D. P. Resler, R. C. Sharp, and E. A. Watson, entitled “Optical phased array technology,” Proceedings of IEEE, Vol. 84, No. 2, pp. 268-298, 1996. This typically consists of a liquid crystal cell with one-dimensional pattern of transparent conductor strips in which each strip defines an element of the linear array. For beam steering in both the azimuth and elevation directions, two such liquid crystal cells are arranged in orthogonal orientations. liquid crystal based optical phased array is a fairly mature technology with very low power consumption due to the capacitive nature of the liquid crystal.
However, a main limitation of the liquid crystal based optical phased array is its slow steering speed (10's of millisecond range) which is due to the slow response time of the liquid crystal-based phase shifting elements. Another disadvantage of liquid crystals is that they operate over a limited temperature range. At low temperatures (<20° C.) the liquid crystal response time significantly degrades due to its increased viscosity, while at higher temperatures (>50° C.) the liquid crystal becomes isotropic and hence loses its functionality. Consequently, for practical purposes, the operating temperature of the liquid crystal based optical phased array should be externally controlled, which further adds to its complexity.
Another problem with the liquid crystal based optical phased array is the presence of grating lobes in the steered beam, which not only severely reduces the optical efficiency of the phased array but also requires complex signal detection circuitry when used in a receiver. In order to eliminate grating lobes in phased arrays, the spacing between array elements must be smaller than the wavelength of the steered beam. In liquid crystal based optical phased arrays, the array element (strip) widths are in the range of 5 to 10 μm, and hence larger than the wavelength of the optical beam in the UV, visible, NIR, and MWIR regions. Reducing the strip width below 5 μm results in a significant field-fringing effect and correspondingly reduces the electro-optic efficiency of the device, since the thickness of the liquid crystal cell is about 4 μm.
There exist prior techniques that can achieve faster scan speeds than liquid crystals, which involve phase modulation via an electrically or thermally controlled index change in semiconductors or electro-optic polymers. For example, one beam steering approach is based on the use of integrated AlGaAs waveguide arrays on a GaAs substrate in which each array element is a tunable phase shifter, such as described in the document from F. Vasey, F. K. Reinhart, R. Houdré, and J. M. Stauffer, entitled “Spatial optical beam steering with an AlGaAs integrated phased array,” Appl. Opt. 32, pp. 3220-3232, 1993.
FIG. 1 illustrates a known optical phased array 10 using optical-waveguide devices to accomplish phase shifting and amplitude adjustment of light from a light source 12, such as disclosed in: “T. E. Dilon, C. A. Schuetz, R. D. Martin, D. G. Mackrides, P. F. Curt, J. Bonnet, and D. W. Prather, “Non-mechanical Beam Steering Using Optical Phased Arrays,” Proc. of SPIE, Vol. 8184, 81840F, pp. 1-11, 2011”.
A linear array of optical-waveguide phase shifters 20 placed side by side provides phase shifts of light from light source 12, after it is split by an optical splitter 22. Optical waveguide phase shifters achieve a 0 to 2π range of phase shift in one phase shifter by controlling a voltage or current applied to the electrodes of the phase shifter.
A series of high output power optical amplifiers 30 is provided to adjust the amplitude of the output of each phase shifter 20. Flexible optical fibers 40 connect each output of amplifiers 30 to an optical emitter output aperture 42 of an emitter surface 44 that comprises a desired two dimensional pattern of emitter output apertures.
FIG. 2 illustrates a front view of an emitter surface 44 having a desired two dimensional pattern of emitter output apertures 42 as illustrated in FIG. 1. Emitter output aperture 42 are arranged along a non-uniform spacing on surface 34 to eliminate the dominant grating lobes and reduce the larger side lobes of the phased array.
An important limitation of using optical fibers is that the minimum spacing between adjacent emitter is very large and thus the average side lobe power is fairly high. Also, since the size of the optical mode in a fiber is large compared to the wavelength of the light, the field of regard is very small (approximately 1 degree). For at least the above reasons, achieving a two-dimensional optical phased array using one or more mono-dimensional arrays of phase shifters and capable of steering an output optical beam over a large field of regard (e.g., ±45-60°) along two orthogonal axes has been difficult.
Another deficiency of known phased arrays such as phased array 10 is that the phase of the light output from the fibers 40, 48 can fluctuate as a result of mechanical vibrations or changes in ambient temperature, thus detrimentally affecting the operation of optical phased array 10.
A two-dimensional optical phased array 71 that consists of a stack of one-dimensional optical phased array slices 72 is illustrated in FIG. 3A. Such two-dimensional optical phased array is described in an article by A. Hosseini, D. Kwong, Y. Zhao, Y.-S. Chen, F. Crnogorac, R. F. W. Pease, and R. T. Chen, “Unequally spaced waveguide arrays for silicon nanomembrane-based efficient large angle optical beam steering,” IEEE J. Selected Topics in Quantum Electronics, Vol. 15, No. 5, pp. 1439-1446, 2009. A 1D optical phased array with non-uniform emitter spacing uses thermo-optic phase shifters fabricated in silicon waveguides is for example described in the document from D. Kwong, A. Hosseini, Y. Zhang, and R. T. Chen, entitled “1×12 Unequally spaced waveguide array for actively tuned optical phased array on a silicon nanomembrane,” Applied Physics Letters, Vol. 99, 051104, 2011.
Each slice 72 of the array consists of a silicon nanomembrane, with a suggested approach for constructing an assembly of multiple slices being a nanomembrane transfer technique described in a document by W. Peng, et al., entitled “Single-crystal silicon/silicon dioxide multilayer heterostructures based on nanomembrane transfer”, Applied Physics Letters, Vol. 90, pp. 183107-1-183107-3, 2007. As described in the article by Peng, et al., a silicon nanomembrane is formed by etching the silicon dioxide layer of a silicon-on-insulator chip, then transfering the thin silicon membrane layer to form a stack of multiple nanomembrane layers. Spin-on-glass can be deposited onto each of the transferred nanomembrane layers to separate that layer from a subsequent nanomembrane layer. Each silicon nanomembrane layer can have a thickness of approximately 100 nm. Optical phased array elements 73 can be formed on the nanomembrane layer, with control contacts 74 arranged on the edge of the nanomembrane layer.
FIG. 3B illustrates a known two-dimensional RF phased array assembly 60 having a two-dimensional array 62 of RF emitters 64. An assembly such as 60 is described in an article by L. Schulwitz and A. Mortazawi, “A Tray Based Rotman Lens Array with Beamforming in Two Dimensions for Millimter-Wave Radar,” IEEE Intl. Symposium on Phased Array Systems and Technology (ARRAY), pp. 850-853, October 2010”. Assembly 60 comprises a plurality of mono-dimensional phased array boards 66 that are placed on top of each other. In assembly 60, the boards 66 are secured at one end to a first mechanical structure supporting array 62 and at an opposite end to a second mechanical structure 68. Array 62 and structure 68 are secured to a base 69. Each board 66 comprises a Rotman lens 70 with a mono-dimensional array of outputs (not shown) connected each to an emitter horn antenna 64 of array 62. The electrical connections to each of the boards 66 are made by means of wires (not shown) that are connected to the side edges of boards 66.
A phased array comprised of emitters 64 arranged regularly such as array 62 of FIG. 3B have grating lobes at angles θgl given by:
            sin      ⁡              (                  θ                      g            ⁢                                                  ⁢            l                          )              -          sin      ⁢                          ⁢              (        θ        )              =            n      ⁢                          ⁢      λ        d  where θ is the scan angle, n is an integer, λ is the wavelength, and d is the array spacing. Typical RF phased arrays have element spacing on the order of λ/2, which enables scanning over a 180° field of regard without the formation of grating lobes.
However, it is difficult to fabricate an optical phased array having a λ/2 element spacing, especially for phased arrays that operate overall the SWIR, MWIR1, MWIR2 or LWIR wavelength bands. For example, the spacing of the emitters of an array as shown in FIG. 3A can be smaller than λ/2 for LWIR light along an x-axis of the array but larger than λ along the y-axis. Also, for an array having emitters emitting lights of more than one wavelength, an emitter spacing smaller than λ/2 for a first light (e.g. LWIR) can be larger than λ for a second light (e.g. SWIR light). As an illustrative example, consider a multi-band array that is designed for a minimum element spacing at the longest wavelength λmax (e.g., dmin=λmax/2=5 μm for the case of λmax=10 μm wavelength). The fixed minimum physical spacing between emitters can correspond to a distance of multiple wavelengths for the shortest wavelength λmin of operation (e.g., dmin=5 μm=3.3·λmin for the case of λmin=1.5 μm). For an element spacing of 2λ, the first grating lobes occur at ±30° from the main beam. Assuming a Gaussian beam, the beam divergence for a 2λ aperture is ±32°, and therefore, the grating lobe will not be cancelled by the element factor. This problem is exaggerated by scanning. For the above example with the main beam pointed at 5°, there is a grating lobe at −25°. Accordingly, a different approach is required to deal with the grating lobes of an optical phased array.
RF phased arrays with non-uniform element spacing have been studied extensively, for example in the document from B. D. Steinberg entitled: “Principles of Aperture and Array System Design”, Wiley-Interscience, New York, 1976; or in the document from C. A. Balanis entitled: “Antenna Theory: Analysis and Design”, 2nd edition, Wiley, New York, Sections 3.6, 6.10, 12.2, and 12.3, 1997; or in the document from R. J. Mailloux entitled “Phased Array Antenna Handbook”, Artech House, Boston, 1994.
However, there exists a need for a two-dimensional optical phased array that produces beams that are free of grating lobes and that have low peak and average sidelobe levels for all emitted wavelengths. There also exists a need for a method for manufacturing such a two-dimensional optical phased array.